Hi, as you can see in this notebook: Drone images segmentation | Kaggle, mainly obtained by copying and pasting from this tutorial: Fine-Tune a Semantic Segmentation Model with a Custom Dataset
after 100 steps in training, mean Iou begins to drop again, while mean accuracy and overall accuracy are almost constant during training and dont grow . What hyperparameters should I put in for better results? Thank you
Hi,
I recommend checking out the hyperparameters of the original paper. From that paper:
We trained the models using AdamW optimizer for 160K iterations on ADE20K, Cityscapes, and 80K iterations on COCO-Stuff. (âŚ) We used a batch size of 16 for ADE20K and COCO-Stuff, and a batch size of 8 for Cityscapes. The learning rate was set to an initial value of 0.00006 and then used a âpolyâ LR schedule with factor 1.0 by default.
Actually with these hyperparameters :
epochs=2
lr = 0.00006
batch_size = 8
I got better results:
(score: 2.932823657989502).
TrainOutput(global_step=80, training_loss=3.012317517399788, metrics={âtrain_runtimeâ: 5060.615, âtrain_samples_per_secondâ: 0.126, âtrain_steps_per_secondâ: 0.016, âtotal_flosâ: 1.123496604205056e+16, âtrain_lossâ: 3.012317517399788, âepochâ: 2.0})